Institute for Systems Research

We begin by quantifying the amount of ﲲesidual redundancy inherent in the LSP's of Federal Standard 1016 CELP. This is done by modeling the LSP's as first and second-order Markov chains. Two models for LSP generation are proposed; the first model characterizes the intra-frame correlation exhibited by the LSP's, while the second model captures both intra-frame and inter-frame correlation. By comparing the entropy rates of the models thus constructed with the CELP rates, it is shown that as many as one-third of the LSP bits in every frame of speech are redundant.

We next consider methods by which this residual redundancy can be exploited by an appropriately designed channel decoder. Before transmission, the LSP's are encoded with a forward error control (FEC) code; we consider both block (Reed- Solomon) codes and convolutional codes. Soft-decision decoders that exploit the residual redundancy in the LSP's are implemented assuming additive white Gaussian noise (AWGN) and independent Rayleigh fading environments. Simulation results employing binary phaseshift keying (BPSK) indicate coding gains of 2 to 5 dB over soft-decision decoders that do not exploit the residual redundancy.

We begin by quantifying the amount of ﲲesidual redundancy inherent in the LSP's of Federal Standard 1016 CELP. This is done by modeling the LSP's as first-and second-order Markov chains. Two models for LSP generation are proposed; the first model characterizes the intra-frame correlation exhibited by the LSP's, while the second model captures both intra-frame and inter-frame correlation. By comparing the entropy rates of the models thus constructed with the CELP rates, it is shown that as many as one-third of the LSP bits in every frame of speech are redundant.

We next consider methods by which this residual redundancy can be exploited by an appropriately designed channel decoder. Before transmission, the LSP's are encoded with a forward error control (FEC) code; we consider both (Reed-Solomon) codes and convolutional codes. Soft-decision decoders that exploit the residual redundancy in the LSP's are implemented assuming additive white Gaussian noise (AWGN) and independent Rayleigh fading environments. Simulation results employing binary phase-shifting keying (BPSK) indicate coding gains of 2 to 5 dB over soft-decision decoders that do not exploit the residual redundancy.

We First derive a maximum likelihood (ML) decoding algorithm for the channel; it turns out that ML decoding is equivalent to decoding a received vector onto either the closest codeword or the codeword that is farthest away, depending on whether an ﲡpparent epidemic has occurred. We next show that the Poly-contagion channel is an ﲡveraged channel in the sense of Ahlswede (and others) and that its capacity is zero. We then demonstrate that the Poly- contagion channel is a counter-example to the adage, ﲭemory cannot decrease capacity ; the capacity of the Poly-contagion channel is actually less than that of the associated memoryless channel. Finally, we consider a finite-memory version of the Poly-contagion model; this channel is (unlike the original) ergodic with a non-zero capacity that increases with increasing memory.

We investigate the problem of detecting a binary iid non-uniform source transmitted across the Markov channel. Two maximum a posteriori (MAP) formulations are considered: a sequence MAP detection and an instantaneous MAP detection. The two MAP detection problems are implemented using a modified version of the Viterbi decoding algorithm and a recursive algorithm. Necessary and sufficient conditions under which the sequence MAP detector becomes useless as well as simulation results are presented. A comparison between the performance of the proposed system with that of a (substantially more complex) traditional tandem source-channel coding scheme exhibits a better performance for the proposed scheme at relatively high channel bit error rates.

The same detection problem is then analyzed for the case of a binary symmetric Markov source. Analytical and simulation results show the existence of a "mismatch" between the source and the channel. This mismatch is reduced by the use of a rate-one convolutional encoder. Finally, the detection problem is generalized for the case of a binary non-symmetric Markov source.

We begin by proposing and analyzing a contagion communication channel. A contagion channel is a system in which noise propagates in a way similar to the spread of an infectious disease through a population; each ﲵnfavorable event (i.e., an error) increases the probability of future unfavorable events. A contagion-based model offers an interesting and less complex alternative to other models of channels with memory like the Gilbert-Elliott burst channel. We call the model set forth the Polya-contagion channel - discrete binary communication channel with additive errors modeled according to the famous urn scheme of George Polya for the spread of contagion.

We nest consider discrete channels with arbitrary (not necessarily stationary ergodic) additive noise. Note that such channels need not be memoryless; in general, they have memory. We show that output feedback does not increase the capacity of such channels. The same result is also shown for a larger class of channels to which additive channels belong, the class of discrete symmetric channels with memory. These channels have the property that their inf-information rate is maximized for equally likely iid input processes.

Finally, we impose average cost constraints on the input of the additive channels, rendering them non-symmetric. We demonstrate that in the case where the additive noise is a binary stationary mixing Markov process, output feedback can increase the capacity-cost function of these channels.